energies Article
Explicit Multipole Formulas for Calculating Thermal Resistance of Single U-Tube Ground Heat Exchangers † Johan Claesson 1 and Saqib Javed 2, * 1 2
* †
ID
Building Physics, Lund University, Lund 221 00, Sweden;
[email protected] Building Services Engineering, Chalmers Technical University, Gothenburg 412 96, Sweden Correspondence:
[email protected]; Tel.: +46-31-772-1155 Presented at the IGSHPA Technical/Research Conference and Expo, Denver, CO, USA, 14–16 March 2017.
Received: 30 November 2017; Accepted: 10 January 2018; Published: 16 January 2018
Abstract: Borehole thermal resistance is both an important design parameter and a key performance characteristic of a ground heat exchanger. Another quantity that is particularly important for ground heat exchangers is the internal thermal resistance between the heat exchanger pipes. Both these resistances can be calculated to a high degree of accuracy by means of the well-known multipole method. However, the multipole method has a fairly intricate mathematical algorithm and is thus not trivial to implement. Consequently, there is considerable interest in developing explicit formulas for calculating borehole resistances. This paper presents derivation and solutions of newly derived second-order and higher-order multipole formulas for calculating borehole thermal resistance and total internal thermal resistance of single U-tube ground heat exchangers. A new and simple form of the first-order multipole formula is also presented. The accuracy of the presented formulas is established by comparing them to the original multipole method. The superiority of the new higher-order multipole formulas over the existing formulas is also demonstrated. Keywords: borehole thermal resistance; internal thermal resistance; multipole method; ground source heat pump (GSHP) systems; boreholes; calculation; second-order; third-order
1. Introduction The use of ground source heat pump (GSHP) systems to provide heating and cooling in buildings has increased at a rapid rate in the last two decades or so. Stimulated by energy prices, technology advances and environmental concerns, the thermal energy utilized by these systems has increased from approximately 500 GWh in 1995 to over 91,000 GWh in 2015. During the same period, the worldwide installed capacities of these systems have increased from under 2000 MWt in 1995 to over 50,000 MWt in 2015 [1]. Comparable growth levels are expected for the future. A typical GSHP system consists of a heat pump, a ground heat exchanger, and auxiliary systems for storage and distribution of thermal energy. The sizing and performance of a heat pump system depends upon the coefficient of performance of the heat pump, which in turn is a function of heat source and heat sink temperatures. Compared to ambient air, ground, in general, is a far superior source or sink of thermal energy because of its relatively stable temperature levels over the year. In a GSHP system, a ground heat exchanger is used to transfer heat to or from the ground. The ground heat exchanger can be of open or closed type. In an open system groundwater is directly used as the heat carrier fluid, whereas in a closed system the heat carrier fluid is circulated in a closed loop, which can be horizontal or vertical. Various heat exchanger configurations can be used in closed-loop vertical systems, including single or double U-tubes, and simple or complex coaxial pipes. Among all types, a borehole heat exchanger with a single U-tube is by far the most commonly used ground heat exchanger in practice because of its low cost, small space requirements, and ease of installation. The scope of this paper is also limited to the application of single U-tubes in borehole heat exchangers. Energies 2018, 11, 214; doi:10.3390/en11010214
www.mdpi.com/journal/energies
Energies 2018, 11, 214
2 of 17
Ground thermal conductivity (λ) and borehole thermal resistance (Rb ) are the two principal parameters that govern the heat transfer mechanism of a borehole heat exchanger, and thus influence the sizing and performance of the overall GSHP system [2]. The heat transfer outside the borehole boundary is dictated by the thermal conductivity of the ground, whereas the heat transfer inside the borehole is characterized by the borehole thermal resistance between the heat carrier fluid and the borehole wall. A high ground thermal conductivity is beneficial for the ground heat transfer. However, being an intrinsic property of the ground, ground thermal conductivity cannot be controlled in practice. On the other hand, a low borehole thermal resistance is desirable for better heat transfer inside the borehole heat exchanger. The borehole thermal resistance depends upon the physical arrangement and the thermal properties of borehole components including grouting, heat exchanger pipes, and the heat carrier fluid. Its value can be engineered to a certain extent by optimizing the geometry and layout of the borehole heat exchanger and by choosing appropriate materials for the borehole components. For any given borehole configuration, the borehole thermal resistance can either be estimated theoretically [3] or measured experimentally [4]. The theoretical methods to estimate borehole thermal resistance include analytical [5–7] or empirical [8–11] formulas based on one-dimensional or two-dimensional steady-state conductive heat transfer in the borehole. As demonstrated by [12,13], there exists a great disparity between borehole thermal resistance values calculated from different theoretical methods. Many of the simplified methods work well in certain situations and poorly in others. The multipole method is an analytical method to determine thermal resistances for any number of arbitrarily placed pipes in a composite region. It is based on two-dimensional steady-state conductive heat transfer in a borehole, and uses a combination of line heat sources and so-called multipoles to determine thermal resistances for any number of arbitrarily placed pipes in a composite region. The method was originally developed by [14,15], and was later revised by [16]. The main difference between the versions of [14] and [16] was in regards to prescribing a constant temperature condition at a suitable radial distance outside the borehole. In the later version, this condition was replaced by a condition on the mean temperature around the borehole wall, which simplified the algorithm. The multipole method has been compared to and tested against several numerical methods. Young [17] compared the tenth-order multipole method with a boundary-fitted coordinates finite volume methods for 18 combinations of shank spacing and grout conductivities, and observed a maximum difference of 0.26% between the two methods. Lamarche et al. [12] calculated a root-mean-square difference of 0.003 between the first-order multipole method and a finite element numerical model for 72 studied cases with varying grout and soil conductivities, and pipe sizes and positing. He [18] studied four cases of different borehole diameters and grout conductivities using a three-dimensional numerical model and the multipole method, and reported a maximum difference of 0.19%. Liao et al. [11] compared the first-order multipole method with a two-dimensional numerical method and noted a maximum relative difference of 0.2% for 18 combinations of shank spacing and grout conductivities. Al-Chalabi [19] studied 12 different cases of pipe positioning, and grout and soil conductivities, using the first-order multipole method and a two-dimensional finite element method, and observed a maximum difference of 1.3% between the two methods. The accuracy and the complexity of the multipole method increases with the number of multipoles used for the calculation. When implemented in a computer program, the order of the multipoles to be used for a calculation is typically prescribed to ten, which was the maximum possible order in the original implementation [14] of the multipole method. Popular ground heat exchanger programs Earth Energy Designer (EED, v3.2. BLOCON: Lund, Sweden) [20] and Ground Loop Heat Exchanger Professional (GLHEPro) [21] also use tenth-order multipoles when calculating the borehole thermal resistance. The tenth-order multipole calculations have an accuracy of over eight decimal digits [22]. On the adverse side, the actual multipole method has a quite rigorous mathematical formulation and a fairly complex algorithm. Its implementation in computer programs requires a considerable amount of coding (the original implementation [14] in FORTRAN was nearly 600 lines in length).
Energies 2018, 11, 214
3 of 17
However, it is possible to simplify the multipole method to explicit closed-form formulas assuming that heat exchanger pipes are placed symmetrically about the center of the borehole. In reality, heat exchanger pipes may be located anywhere in the borehole as long as they do not overlap each other. The position of pipes also varies along the depth of the borehole. However, in the absence of any a priori knowledge of the pipe positions, it is reasonable to assume that the heat exchanger pipes are symmetrically placed about the center of the borehole. So far, closed-form multipole formulas for zeroth-order and first-order have been developed for the case of a single U-tube with symmetrical pipes [23]. In this paper, we present derivation and solutions of new second-order and higher-order multipole formulas for calculating borehole thermal resistance of single U-tube ground heat exchangers with symmetric pipes. The presented formulas are tested against the original multipole method (i.e., the tenth-order multipole calculation) as well as previously-derived zeroth-order and first-order multipole formulas. 2. Borehole Thermal Resistance The borehole thermal resistance (Rb ) is the ratio of the temperature difference between the heat carrier fluid and the borehole wall to the heat transfer rate per unit length of the borehole. It is defined locally at a specific depth in the borehole as represented in Equation (1), where Tf,loc is the local mean fluid temperature, Tb is the average borehole wall temperature and qb is the heat transfer rate per unit length of the borehole. Tf,loc − Tb (1) Rb = qb Borehole thermal resistance is sometimes treated as a sum of resistances in series. The total borehole resistance (Rb ) i.e., fluid-to-ground resistance is considered to be made up of two major parts: pipe resistance (Rp ) and grout resistance (Rg ). The pipe resistance includes both conductive resistance of the pipe (Rpc ) and convective resistance of the fluid (Rpic ). The grout resistance constitutes the thermal resistance between the outer pipe wall of the U-tube and the borehole wall. Its value depends on the pipe resistance and the ground thermal conductivity [13]. As the individual pipe resistances are in parallel with each other, the pipe resistance is generally calculated for a single pipe and is then divided by the total number (N) of pipes. Rb = Rg +
Rpc + Rpic Rp = Rg + N N
(2)
Nevertheless, in order to fully understand the concept of borehole thermal resistance, it is helpful to use a resistance network. Figure 1 shows an example of such a network for a single U-tube borehole. Other networks forms can also be formulated, e.g., see [11,24] but any such representation is an Energies 2018, 11, 214 4 of 17 approximation to reality under network-specific assumptions and restrictions.
Tb,av
Figure 1. Δ resistance network for a single U-tube in a borehole. Figure 1. ∆ resistance network for a single U-tube in a borehole.
The three resistance Δ network of Figure 1 is based on the following relations between heat flows q1 and q2 and fluid temperatures Tf1 and Tf2.
q1 =
Tf1 -Tb,av R1b
+
Tf1 -Tf2 R12
(3)
Energies 2018, 11, 214
4 of 17
The three resistance ∆ network of Figure 1 is based on the following relations between heat flows q1 and q2 and fluid temperatures Tf1 and Tf2 . q1 =
Tf1 − Tb,av T − Tf2 + f1 R1b R12
(3)
q2 =
Tf2 − Tb,av T − Tf1 + f2 R2b R12
(4)
In the above equations R1b and R2b are thermal resistances between pipe 1 or 2 and the borehole wall, respectively, and R12 is the fluid-to-fluid thermal resistance between pipes 1 and 2. These resistances include the thermal resistances of the fluid and the pipe. Using the ∆ network of Figure 1, the total borehole resistance, Rb , can be determined by setting the fluid temperatures Tf1 = Tf2 and solving for two parallel resistances R1b and R2b : Rb =
R1b R2b R1b + R2b
(5)
The ∆ network can also be used to determine resistance Ra by setting the heat flows q1 = −q2 . The resistance, Ra , is the total internal resistance between the upward and downward flowing legs of the U-tube when there is no net heat flow from the borehole. It is related to the fluid-to-fluid internal resistance R12 by the thermal network of Figure 1. The resistance R12 , which is sometimes also referred to as the direct coupling resistance, is the effect of the network representation and is not a directly measurable physical resistance. Resistance R12 is in parallel with the series resistances R1b and R2b . Ra =
R12 ( R1b + R2b ) R12 + R1b + R2b
(6)
The concept of total internal resistance Ra is critical to the understanding of the effective borehole thermal resistance Rb∗ , which is the thermal resistance between the heat carrier fluid, characterized by a simple mean Tf of the inlet and outlet temperatures, and the average borehole wall temperature Tb . The effective borehole thermal resistance is mathematically expressed as Equation (7). Rb∗ =
Tf − Tb qb
(7)
The difference between borehole thermal resistance Rb , defined by Equation (1), and effective borehole thermal resistance Rb∗ , defined by Equation (7), is that the borehole thermal resistance applies locally at a specific depth in the borehole, whereas effective borehole thermal resistance applies to the entire borehole. The effective borehole thermal resistance is what is measured in an experiment, for example an in-situ thermal response test. The experimental measurements of mean fluid temperature taken at the top of the borehole include the effects of thermal short-circuiting between the upward and downward flows in the U-tube. Consequently, the effective borehole thermal resistance is often higher than the borehole thermal resistance due to the higher fluid temperature caused by the thermal short-circuiting between the U-tube pipes. It is possible to calculate the effective borehole thermal resistance from the borehole thermal resistance [23,25,26]. Equations (8) and (9), developed by [23], are commonly used for this purpose. These equations correspond to two distinct boundary conditions of uniform borehole wall temperature and uniform heat flux along the borehole, respectively. Both these are limiting conditions and the real situation falls somewhere in between. Therefore, the effective borehole thermal resistance is commonly determined as the mean value between these two equations. Calculation of the effective borehole thermal resistance from Equations (8) and (9) requires knowledge of total internal thermal resistance
and uniform heat flux along the borehole, respectively. Both these are limiting conditions and the real situation falls somewhere in between. Therefore, the effective borehole thermal resistance is commonly determined as the mean value between these two equations. Calculation of the effective borehole thermal resistance from Equations (8) and (9) requires knowledge of total internal thermal resistance a and Energies 2018,R11, 214 direct coupling resistance R12 between the two U-tube legs, respectively, in addition 5 of 17 to the borehole thermal resistance Rb. 2
Ra and direct coupling resistance R12 between the two 1 æçU-tube H ö÷÷ legs, respectively, in addition to the çç (8) Rb* = Rb + ÷÷ borehole thermal resistance Rb . ÷2 3R a çè rfcV f f ø 1 H Rb∗ = Rb + (8) 3Ra ρf cf Vf 4R H 1s Rb* = Rb h ⋅ coth h, h= ⋅ 1+ b (9) 4R Hr c V 12R Rb12 f f·f b 1+ Rb∗ = Rb η · cothη, η= (9) ρf cf Vf 2Rb R12
3. Multipole Multipole Method Method for for N N Pipes in a Borehole Borehole 3. The general general multipole multipole method may be applied for any number, N, N, of of pipes pipes in in the the borehole borehole as as The shown in Figure 2. The center of pipe n lies at (x n ,y n ). The thermal conductivity of the grout region shown in Figure 2. The center of pipe n lies at (xn ,yn grout region outside the the pipes pipesisisλλbb (W/m-K), (W/m-K), and λ. The borehole borehole radius radius is is outside and of of the the ground ground outside the boreholes is λ. pipes is is rpp. The pipe to to the the grout grout rrbb and the outer radius of the pipes The thermal thermal resistance from the fluid in the pipe adjacent to to the the pipe pipe is is RRpp (m-K/W). (m-K/W). adjacent
y Tb,av
q1 qn Tf1
(xn,yn) Tfn
Rp
x
qN λb
rpn
TfN
λ rb Figure heat conduction in aa composite Figure 2. 2. Steady-state Steady-state heat conduction in composite circular circular region region with with heat heat flows flows between between N N number of pipes and the adjacent ground region. number of pipes and the adjacent ground region.
The (steady-state) (steady-state)temperature temperature field T(x,y) is be to determined be determined for set any of prescribed heat The field T(x,y) is to for any ofset prescribed heat fluxes qn,. .n. ,=N,1,and … N, any prescribed average temperature Tb,avthe around the wall. borehole wall. qfluxes anyand prescribed average temperature Tb,av around borehole The fluid n , n = 1, The fluid temperature n From = 1, …, N.the From this the borehole thermalare resistances are temperature in pipe n isin Tfnpipe , n =n1,is. .T. fn, ,N. this borehole thermal resistances determined. determined. The temperature field consists of a linear combination of line heat source solutions and The temperature field consists of a linear combination of line heat source solutions and so-called so-called multipole solutions at the center of each pipe: multipole solutions at the center of each pipe: N
qn T ( x, y) = Tb,av + ∑ · Re[W0 (z, zn )] + Re 2πλ b n =1
"
N
J
∑ ∑ Pn,j · Wj (z, zn )
# (10)
n =1 j =1
The following complex notations are used: z = x + i · y,
z = x − i · y,
zn = xn + i · yn ,
n = 1, . . . , N.
(11)
The real part of the function W 0 (z,zn ) gives the temperature field for a line heat source at z = zn with the strength qn = 2πλb (W/m):
W0 (z, zn ) =
ln
rb z−zn
rb2 rb2 −z·zn
+ σ · ln (1 + σ ) · ln rb z−zn + σ ·
1+ σ 1− σ
| z | ≤ rb · ln
rb z
| z | ≥ rb
(12)
Energies 2018, 11, 214
6 of 17
The upper expression of Equation (12) is valid in the borehole outside the pipes and the lower one in the ground outside the borehole. The first top logarithm represents an ordinary line heat source, while the second term represents a line source at the mirror point z = rb2 /zb . The two expressions, and the ensuing radial heat flux at the grout and ground side, are constructed to become equal at the borehole radius |z| = rb . This means that the solution satisfies the internal boundary conditions at the borehole wall. The average temperature Tb,av around the borehole wall is zero. The last term in Equation (10) for the temperature field involves a multipole solution of order j at z = zn : j j rp · z rp + σ · r2 −z·z | z | ≤ rb z−zn n Wj (z, zn ) = j = 1, 2, . . . (13) j b (1 + σ ) rp z ≥ r | | b z−zn The above solutions satisfy the internal boundary conditions of continuous temperature and radial heat flux at the borehole wall. The average temperature at the borehole wall is zero. Multipoles up to order j = J are used. The first top term in the above expressions, in local polar coordinates around the outer periphery of pipe n, becomes:
rp z−zn
j
= z − z n = ρ · ei ·ψ =
j
rp ρj
· e −i · ψ · j =
ρ = rp = cos( j · ψ) − i · sin( j · ψ),
−π < ψ ≤ π.
(14)
The multipole strength factors Pn,j are complex-valued numbers. This means the expression (10) can account for a Fourier expansion of any temperatures around the pipes with Cosine and Sine terms up to order j = J. The boundary condition at the periphery of pipe n involves a balance between the radial heat flux and the temperature difference between fluid temperature Tfn and the temperature outside the pipe, which varies around the pipe. Let as in Equation (14), ρ be the radial distance from the center of pipe n and ψ the polar angle. The temperature in the grout at the pipe wall and the radial heat flux multiplied by the pipe resistance per unit pipe area are given by the two left-hand terms below: i ∂ h efn (ψ) ' Tfn . =T T ( x, y)|ρ,ψ T ( x, y)|ρ=rp ,ψ + Rp · 2πrp · (−λb ) · | {z } ∂ρ ρ =rp
(15)
− β ·rp
The left-hand expression will depend on the polar angle ψ. The exact boundary condition is that the expression becomes constant and equal to the required fluid temperature Tfn to obtain the prescribed heat fluxes qn . The temperature field of Equation (10), taken for any set of multipole factors Pn,j , is inserted in Equation (15). It is possible by quite lengthy calculation involving complex-valued formulations and Taylor expansions to separate the expressions in Fourier exponentials of any order j. The Fourier coefficient of order j for any pipe n depends on the heat fluxes and multipole factors: " e Tfn (ψ) = Re Fn,0 (q, P) +
∞
∑ Fn,j (q, P) · e
# i·ψ· j
.
(16)
j =1
Here, the vector q represents the N prescribed heat fluxes qn , and the matrix P represents the N·J complex-valued multipole factors Pn,j . The multipole factors are now chosen so that the Fourier coefficients up to order J become zero. This gives N·J equations for the multipole factors: Fn,j (q, P) = 0,
n = 1, . . . , N,
j = 1, . . . , J
→
Pn,j
(17)
Here, the vector q represents the N prescribed heat fluxes qn, and the matrix P represents the N·J complex-valued multipole factors Pn,j. The multipole factors are now chosen so that the Fourier coefficients up to order J become zero. This gives N·J equations for the multipole factors: Energies 2018, 11, 214
Fn , j (q, P) = 0,
n = 1, ... N ,
j = 1, ... J
Pn , j
7 of 17 (17)
The equations are complex-valued with a real and an imaginary part. This means that the The of equations are complex-valued with a real imaginary This meansterm that the number number unknowns and equations are 2·N·J forand the an general case. part. The zero-order in Equation of unknowns and equations are 2 · N · J for the general case. The zero-order term in Equation (16) determines the required fluid temperature Tfn, and the terms above j = J represent the error(16) for determines the required Tfn , and thepolar termsangle above the heat balance over thefluid pipe temperature wall as a function of the ψ:j = J represent the error for the heat balance over the pipe wall as a function of the polar angle ψ:
é ¥ ù (y) -T ="Re ê å F (q, P) ⋅ ei⋅y⋅j#ú . Tfn = Re éêF0,n (q, P)ùú , T ∞ê fn fn k ,n ú ë û êë j =J +F1k,n (q, P) · ei·ψ· j úû . efn (ψ) − Tfn = Re ∑ Tfn = Re[ F0,n (q, P)], T
(18) (18)
j = J +1
4. Two Symmetrically Placed Pipes in a Borehole 4. Two Symmetrically Placed Pipes in a Borehole Figure 3 shows the case that is considered in this study. There are two symmetrically placed Figure 3 shows the case that is considered in this study. There are two symmetrically placed pipes pipes in a grouted borehole. The center of the two pipes lie at (xp,0) and (−xp,0), respectively. in a grouted borehole. The center of the two pipes lie at (xp ,0) and (−xp ,0), respectively. The prescribed The prescribed heat fluxes from the pipes are q1 and q2 (W/m). The corresponding fluid temperatures heat fluxes from the pipes are q1 and q2 (W/m). The corresponding fluid temperatures are Tf1 and Tf2 . are Tf1 and Tf2. The main objective is to present a very precise method to calculate the relations The main objective is to present a very precise method to calculate the relations between heat flows between heat flows and fluid temperatures. and fluid temperatures.
y Tb,av q2
Rp
q1
rp
Tf2
x
Tf1 -xp
xp
λb
λ
rb Figure Figure 3. 3. Two Two symmetrically symmetrically placed placed pipes pipesin inaagrouted groutedborehole. borehole.
The primary primary input input data dataare: are: The l ,
l,
λb , b λ,
r ,
x ,
r ,
R ,
rb , b xp ,p rp ,p Rp ,p
q ,
q1 ,1
q ,
q2 , 2 J,
J,
T
Tb,avb, .av
.
(19) (19)
The thermal resistance may be represented by a dimensionless parameter β and the ratio between the two conductivities by the parameter σ. The circles of the pipes and the borehole should not intersect. This gives: λ −λ rp ≤ xp ≤ rb − rp ; β = 2πλb Rp , σ= b . (20) λb + λ The number of equations and unknowns for multipole factors for two pipes are in general 2·2·J. But in the considered symmetrical case, the multipole factors Pn,j turn out to be real-valued numbers. The number of unknowns and equations are reduced to 2·J for the two sets of multipoles P1,j and P2,j . 4.1. Multipole Relations for Even and Odd Solutions The multipoles for the two pipes depend on the prescribed heat fluxes. The temperature field is symmetric in x for equal heat fluxes: Even case : s = +1,
q2 = s · q1 ,
Tb,av = 0
⇒
T ( x, y) = T (− x, y).
(21)
Energies 2018, 11, 214
8 of 17
The temperature field is odd in x for antisymmetric heat fluxes: Odd case : s = −1,
q2 = s · q1 ,
Tb,av = 0
⇒
T ( x, y) = − T (− x, y).
(22)
The average temperature at the borehole wall is chosen to be zero in both cases: Tb,av = 0. There is a set of multipole factors for the even case and another one for the odd case. It is found for these two cases of high symmetry that the multipoles of pipe 2 are closely related to the multipoles of pipe 1 in the even and odd cases. The following relations are valid for the even case s = +1 and for the odd case s = −1: ( + ↔ s=1 j ± ± P2,j = s · (−1) · P1,j , j = 1, . . . , J. (23) − ↔ s = −1, The upper index + represents the even temperature solution and s = +1, while the upper index − represents the odd solution and s = −1. With these notational conventions, the formulas below for the even and odd cases can be presented in the same formulas, where s = ±1 becomes a parameter. The temperature solutions for the even and odd cases determine the required fluid temperatures, − which define an even and an odd thermal resistance R+ J and R J , respectively, in the following way: + + Even case q2 = q1 , Tb,av = 0 : Tf2 = Tf1 ,
+ Tf1 = R+ J · q1 .
− − Odd case q2 = −q1 , Tb,av = 0 : Tf2 = − Tf1 ,
− Tf1 = R− J · q1 .
(24) (25)
These even and odd thermal resistances are obtained from the multipole solutions. They are the required fluid temperatures for unit heat flux. These thermal resistances depend on the choice of the number of multipoles J. It will be shown below that the accuracy increases rapidly with increasing J. 4.2. Formulas for Even and Odd Thermal Resistances The temperature field of Equation (10) may be used without the right-hand sum of multipoles. In this case J = 0, and only the sum of line sources are used. The thermal resistances for the two symmetrically placed pipes then become: R0+
" !# ! rb2 rb4 1 = Rp + · ln + σ · ln 4 , 2πλb 2rp xp rb − xb4
(26)
" !# rb2 + xp2 2xp 1 = Rp + · ln + σ · ln 2 . 2πλb rp rb − xp2
(27)
R0−
The correction to the zero-order resistance for any positive J is given by a formula of the type: ± ± R± J = R0 − B J ,
s = ±1,
J = 1, 2, 3, . . . .
(28)
− The quantities B+ J and B J are obtained from the even and odd solutions for the considered multipole order J. In the formulas below the following auxiliary (dimensionless) parameters are used:
p0 =
rp , 2xp
p1 =
rp xp rb2
−
xp2
,
p2 =
rp xp rb2
+
xp2
,
bk =
1 − kβ , 1 + kβ
k = 1, 2, . . . .
(29)
J = 1, 2, 3, . . . .
(30)
The B-values for any positive J are obtained from the formula: B± J =
tr −1 1 · VJ± · MJ± · VbJ± , 2πλb
s = ±1,
Energies 2018, 11, 214
9 of 17
The larger dot indicates a matrix/vector multiplication. The formula involves two vectors VJ and VbJ , and a matrix MJ for the even case s = +1 and the odd case s = −1: ± M1,1 . MJ± = .. M± J,1
VJ±
tr
=
V1±
· · · VJ±
,
··· .. . ···
± M1,J .. . , M± J,J
b1 · V1± .. . VbJ± = . b J · VJ±
(31)
Note that inverses of the M matrices appear in Equation (30). The components of the vectors VJ and VbJ are given by: Vk± = (− p0 )k · s + σ · p1k + σ · (− p2 )k · s,
Vbk± = bk · Vk± ,
k = 1, . . . , J.
(32)
The M matrices are given by: ( ± Mk,j = k · δk,j + bk · A± k,j ,
k = 1, . . . , J,
j = 1, . . . , J;
δk,j =
1
j=k
0
j 6= k
.
(33)
Here, the two matrices A+ and A− , which account for various interactions between line sources and multipoles in the boundary conditions at the two pipes, are fairly complicated: A± k,j =
( k + j −1) ! (k−1)! ·( j−1)!
· (− p0 )k+ j · s + σ k j ·
A± k,j
=
A± j,k ,
min(k,j)
∑
i =0
s = ±1,
( k + j −1− i ) ! i! ·(k −i )! · ( j−i )!
h i k + j −i · (2p0 )i p1 + (− p2 ) j+k−i · s ,
(34)
k = 1, . . . , J,
j = 1, . . . , J.
It may be noted that the A matrices are symmetric. The matrix formula, Equation (30), becomes a sum over all components of the inverse of the M matrix: B± J
J J −1 1 ± ± = · · b j · Vj± , ∑ Vk · MJ 2πλb k∑ k,j j = 1 =1
J = 1, 2, 3, . . . ,
s = ±1.
(35)
4.3. Explicit Multipole Formulas for First, Second and Third-Order The zero-order resistances R0+ and R0− are given by the explicit formulas of Equations (26) and (27), respectively. Explicit formulas may also be derived for J = 1 and J = 2. The first-order B1± corrections are given by Equation (35) for J = 1, in which case the inverse of the matrix is equal the inverse of the single matrix component: B1± =
V ± · b1 · V1± 1 . · 1 ± 2πλb M1,1
(36)
The quantities in this formula are defined in Equations (32)–(34): V1± = − p0 · s + σ p1 − σp2 · s,
Vb1± = b1 · V1± ,
s = ±1, (37)
± ± M1,1 = 1 + b1 · A1,1 ,
± A1,1 = p20 · s + σ · [ p1 · ( p1 + 2p0 ) + p2 · ( p2 − 2p0 ) · s].
The final formula is then: B1± =
b1 · (− p0 · s + σp1 − σp2 · s)2 1 2 . · 2πλb 1 + b1 · p0 · s + σ · [ p1 · ( p1 + 2p0 ) + p2 · ( p2 − 2p0 ) · s]
(38)
Energies 2018, 11, 214
10 of 17
Here, the five dimensionless parameters are defined by Equations (20) and (29). The second-order B2± corrections are given by Equation (35) for J = 2: B2± =
tr h −1 i 1 · V2± · M2± ·Vb2± . 2πλb
The M matrices and their inverses are: ! ± ± −1 1 + b1 A1,1 b1 A1,2 1 ± · M2 = , M2± = ± ± b2 A1,2 2 + b2 A2,2 det(M2± )
± 2 + b2 A2,2 ± −b2 A1,2
(39)
± −b1 A1,2 ± 1 + b1 A1,1
! . (40)
The first-order components in Equation (37) for J = 1 are still valid. The second-order components of the symmetric A matrices and V vectors are from Equations (32) and (34): ± ± A1,2 = A2,1 = −2 p30 · s + 2σ · p21 · ( p1 + 2p0 ) − p22 · ( p2 − 2p0 ) · s , ± A2,2 = 6p40 · s + 2σ · p21 3p21 + 8p1 p0 + 4p20 + p22 3p22 − 8p2 p0 + 4p20 · s ,
(41)
V2± = p20 · s + σ · p21 + σ · p22 · s. A more explicit formula can now be obtained from Equations (35), (39) and (40): B2± =
1 · 2πλb
b1 V1±
2 2 ± ± ± + b2 V2± · 1 + b1 A1,1 · 2 + b2 A2,2 − 2b1 b2 V1± V2± · A1,2 . 2 ± ± ± 1 + b1 A1,1 · 2 + b2 A2,2 − b1 b2 A1,2
(42)
The third-order B± J corrections for J = 3 involve the inversion of the 3 × 3 M matrices. A more explicit formula cannot be given. The third-order B3± corrections are given by Equation (30) for J = 3: B3± =
tr h −1 i 1 · V3± · M3± ·Vb3± . 2πλb
(43)
Here, Vk± and Vbk± are given by Equation (31) for J = 3. The M matrices have the form: ± 1 + b1 A1,1 ± M3± = b2 A1,2 ± b3 A1,3
± b1 A1,2 ± 2 + b2 A2,2 ± b3 A2,3
± b1 A1,3 ± b2 A2,3 . ± 3 + b3 A3,3
(44)
± Here, A± k,j is from Equation (34) for indices k or j equal to three. The third-order B3 corrections of Equation (43) are given by Equation (45) for J = 3:
B3± =
h 3 3 −1 i 1 · ∑ ∑ Vk± · M3± · b j · Vj± . 2πλb k=1 j=1 k,j
(45)
The formula contains all 9 components of the inverse matrix to M3± . 4.4. Thermal Resistances R1b , R12 , Rb and Ra Final formulas for thermal resistances R1b , R12 , Rb and Ra can now be obtained by again considering the ∆ network of Figure 1. For the case of equal pipes in symmetric positions, the thermal resistances R1b is equal to R2b , and thus the heat fluxes of Equations (3) and (4) become: q1 =
Tf1 − 0 Tf1 − Tf2 + , R1b R12
q2 =
Tf2 − 0 Tf2 − Tf1 + . R1b R12
(46)
Final formulas for thermal resistances R1b, R12, Rb and Ra can now be obtained by again considering the Δ network of Figure 1. For the case of equal pipes in symmetric positions, the thermal resistances R1b is equal to R2b, and thus the heat fluxes of Equations (3) and (4) become:
q1 =
Energies 2018, 11, 214
Tf1 - 0 R1b
+
Tf1 -Tf2 R12
,
q2 =
Tf2 - 0 R1b
+
Tf2 -Tf1 R12
.
11 (46) of 17
Using and (46), (46), the the thermal thermal networks networks for for even even(q(q22 == qq11)) and and odd odd(q(q22 == − −qq11)) Using Equations Equations (24), (24), (25) (25) and cases cases can can be be drawn drawn as as shown shown in in Figure Figure 4. 4. q1−q1= 0
2q1
0
0 R1b
+ Tf1 −Tb,av
+ Tf1 −Tb,av
R12
q1
R1b
R1b
R1b
0
−(Tf1−−Tb,av) −q1
q1
R12 2
Tf1− −Tb,av R12 2
q1
Figure 4. 4. Thermal Thermalresistance resistance networks networks for for equal equal pipes pipes in in symmetric symmetric positions positions for for even even (left) (left) and and odd odd Figure (right) cases. cases. (right)
As As seen seen from from the the even even case case of of Figure Figure 4, 4, the the borehole borehole thermal thermal resistance resistance R Rbb between between the the fluid fluid in in pipes and the borehole wall consists of two parallel equal resistances, each of value R 1b. On the other pipes and the borehole wall consists of two parallel equal resistances, each of value R1b . On the other hand, hand, from from the the odd odd case case of of Figure Figure 4, 4, it it can can be be noticed noticed that that the the total total internal internal resistance resistance R Raa between between the the two pipes consists of a pair of equal series resistances, each of value 0.5 R 12. Hence, Equations (5) and two pipes consists of a pair of equal series resistances, each of value 0.5 R12 . Hence, Equations (5) and (6) (6) can can now now be be simplified simplified to: to:
R (J ) R 1b ( J) RbR ( Jb)(J=) = 1b 2 , , 2
2R (J )R12 (J ) 2R 1b ( J )R ( J ) R aR ( Ja)(J=) = 2R1b (J ) +12R (J ), , 2R1b1b ( J ) + R1212( J )
J = J= 0, 0, 1,1,2,2,3,3,... . . . ..
(47) (47)
Final and R12 are obtained in the form of even and odd Final formulas formulas for for network network resistances resistances R R1b 1b and R12 are obtained in the form of even and odd + − thermal resistance resistance RRJ+ and using Figure 4 and Equations (24)(24) andand (25). andR JRusing Figure 4 and Equations (25). thermal J
R1b ( J ) =
R+ J ,
J
1 2 1 − = R ( J) + R ( J) , RJ 12 1b
R12 ( J ) =
− 2R+ J RJ − R+ J − RJ
,
J = 0, 1, 2, . . . ,
(48)
Similarly, final formulas for thermal resistances Rb and Ra are obtained in the form of even and − odd thermal resistance R+ J and R J using Equations (47) and (48). Rb ( J ) =
R+ J 2
,
Ra ( J ) = 2R− J ,
J = 0, 1, 2, 3, . . . .
(49)
− For Equations (48) and (49), the values of thermal resistance R+ J and R J are obtained from Equations (28) and (38) for the first-order multipoles, from Equations (28) and (42) for the second-order multipoles, and from Equations (28) and (45) for the third-order multipoles.
5. Comparison with Existing Multipole Solutions In this section, the new higher-order multipole formulas for borehole thermal resistance and total internal thermal resistance are compared with previously published results. The comparison is made using a reference dataset provided by [13], who compared and benchmarked the borehole thermal resistance estimations from ten different analytical methods against the tenth-order multipole method for 216 different cases. The authors showed that compared to other methods, the zeroth-order and first-order multipole formulas provide better accuracies over the entire range of studied parameters. In this paper, we will also benchmark the newly developed second-order and third-order multipole formulas against the tenth-order multipole method. The second-order and third-order multipole
Energies 2018, 11, 214
12 of 17
formulas will also be compared to each other and to the zeroth-order and first-order multipole formulas to demonstrate improvements in the accuracy of the calculated results. Table 1 provides the detailed summary of the 216 comparison cases provided by the reference dataset used in this paper for the comparison of the second-order and third-order multipole formulas. The cases cover three different borehole diameters of 96 mm, 192 mm, and 288 mm. The U-tube outer pipe diameter value is held fixed at 32 mm for all cases. The total pipe resistance Rp also remains constant at 0.05 m-K/W. For each borehole diameter, three shank spacing configurations, i.e., close, moderate and wide, corresponding, respectively, to Paul’s [8] Configuration A, Configuration B and Configuration C, are considered. Four levels of ground thermal conductivity ranging from 1–4 W/m-K, and six levels of grout thermal conductivity ranging from 0.6–3.6 W/m-K are used. Given the existing and reasonably foreseeable values of design parameters, the 216 cases used for the comparison bracket almost all real-world single U-tube borehole heat exchangers. Table 1. Summary of comparison cases provided by [13].
1 2
Parameters
Levels
No. of Levels
Ratio of borehole diameter to outer pipe diameter 1 , 2rb /2rp
3, 6, 9
3
Shank spacing configuration, 2xp (mm)
Close, Moderate, Wide 2 For rb /rp = 3, 2xp = 32, 43, 64 For rb /rp = 6, 2xp = 32, 75, 160 For rb /rp = 9, 2xp = 32, 107, 256
3
Ground thermal conductivity, λ (W/m-K)
1, 2, 3, 4
4
Pipe outer diameter (2rp ) is fixed at 32 mm, borehole diameters (2rb ) are 96 mm, 192 mm, and 288 mm. Corresponding to Paul’s [8] A–C configurations.
Results of the second-order and third-order multipole formulas for calculating the grout thermal resistance and the total internal thermal resistance are summarized in Tables 2 and 3, respectively. Each entry in these two tables represents the mean or maximum error in percentage for a sample containing two three values of grout thermal conductivity, three values of borehole diameter, and four values of ground thermal conductivity. The errors have been determined by comparing the results of multipole formulas to the tenth-order multipole method. Table 2. Mean and maximum absolute percentage errors in calculation of the grout thermal resistance (Rg ) for the 216 cases provided by [13]. Ground Conductivity Method
Shank Spacing Configuration
Low (0.6–1.2 W/m-K)
Moderate (1.2–2.4 W/m-K)
High (2.4–3.6 W/m-K)
Mean
Max
Mean
Max
Mean
Max
Zeroth-order multipole
Close Moderate Wide
6.1 3.3 8.9
12.4 10.8 30.4
2.9 1.5 1.9
8.0 6.5 11.2
1.0 0.4 0.3
2.5 1.6 1.8
First-order multipole
Close Moderate Wide
0.2 0.0 0.5
0.4 0.2 2.2
0.2 0.0 0.0
0.6 0.2 0.2
0.7 0.1 0.1
1.5 0.5 0.6
Second-order multipole
Close Moderate Wide
0.0 0.0 0.1
0.0 0.0 0.3
0.1 0.0 0.0
0.2 0.1 0.0
0.2 0.0 0.0
0.5 0.1 0.1
Third-order multipole
Close Moderate Wide
0.0 0.0 0.0
0.0 0.0 0.0
0.0 0.0 0.0
0.1 0.0 0.0
0.1 0.0 0.0
0.2 0.0 0.0
Energies 2018, 11, 214
13 of 17
Table 3. Mean and maximum absolute percentage errors in calculation of the total internal thermal resistance (Ra ) for the 216 cases provided by [13]. Ground Conductivity Method
Shank Spacing Configuration
Low (0.6–1.2 W/m-K)
Moderate (1.2–2.4 W/m-K)
High (2.4–3.6 W/m-K)
Mean
Max
Mean
Max
Mean
Max
Zeroth-order multipole
Close Moderate Wide
23.3 1.8 1.9
37.6 7.8 8.5
6.9 1.1 0.4
15.0 6.0 2.3
1.2 0.3 0.2
2.7 1.8 1.3
First-order multipole
Close Moderate Wide
3.2 0.2 0.3
5.9 0.8 1.2
0.6 0.0 0.0
0.7 0.2 0.1
0.7 0.1 0.0
0.7 0.2 0.1
Second-order multipole
Close Moderate Wide
0.4 0.0 0.0
1.0 0.0 0.2
0.2 0.0 0.0
0.2 0.0 0.0
0.1 0.0 0.0
0.2 0.0 0.0
Third-order multipole
Close Moderate Wide
0.0 0.0 0.0
0.1 0.0 0.0
0.1 0.0 0.0
0.1 0.0 0.0
0.0 0.0 0.0
0.0 0.0 0.0
Table 2 shows that the grout thermal resistance (Rg ) values obtained from the second-order and third-order multipole formulas are, respectively, within 0.5% and 0.2% of the tenth-order multipole method for all 216 cases. Also, the mean absolute percentage error of the results obtained from the second-order and third-order multipole formula are, respectively, smaller than 0.2% and 0.1%. In comparison, the mean and maximum absolute percentage errors for the zeroth-order multipole formula are as high as 9% and 30%, respectively. The first-order multipole formula has smaller errors than the zeroth-order formula. The mean and maximum absolute percentage errors for the first-order multipole formula are 0.7% and 2.2%, respectively. Table 3 shows that the total internal thermal resistance (Ra ) values calculated from the second order and third-order multipole formulas are, respectively, within 1% and 0.1% of the tenth-order multipole method for all 216 cases. The mean absolute percentage error of the results obtained from the second-order and third-order multipole formulas never exceed 0.4% and 0.1%, respectively. In comparison, the zeroth-order and the first-order multipole formulas give maximum absolute percentage errors of approximately 38% and 6%, respectively. The mean absolute percentage errors of the zeroth-order and the first-order multipole expressions are as high as 23% and 3%, respectively. As shown and discussed above, compared to the zeroth-order and first-order multipole formulas, the accuracy of the newly developed second-order and third-order multipole formulas to calculate borehole thermal resistance and total internal thermal resistance is higher by several orders of magnitude. Even though the second-order and third-order multipole formulas presented in this paper are more complicated than other analytical expressions including the zeroth-order and first-order formulas, they are still simple enough to apply for computation purposes. The implementation of the second-order multipole formulas involves approximately ten lines of coding of rather compact and simple algebraic expressions. Whereas, the implementation of third-order multipole formulas requires programing of a more complex set of equations involving vector and matrix algebra. Although, both second-order and third-order multipole formulas offer a significant improvement over the original implementation of the multipole method, which required about 600 lines of FORTRAN coding, but the second-order multipole formulas are intrinsically more straightforward to implement, while providing almost similar levels of accuracy as third-order multipole formulas. Figures 5–7 present a selection of representative comparison results of second-order multiple formulas against zeroth-order, first-order and tenth-order multipoles. The results shown in these figures are for a single ground thermal conductivity of 4.0 W/m-K. The left-side figures show the grout thermal resistance Rg
0.20 0.20 0.16 0.16 0.12 0.12 0.08 0.08 0.04 0.04 0.00 0.00 0.0 0.0
0th-order Multipole 1st-order Multipole 0th-order Multipole 2nd-orderMultipole Multipole 1st-order 10th-orderMultipole Multipole 2nd-order 10th-order Multipole
0.6 0.6
1.2 1.2
1.8 2.4 2.4 λ1.8 [W/m-K] g λg [W/m-K]
3.0 3.0
3.6 3.6
Ra [m-K/W] Ra [m-K/W]
Rg [m-K/W] Rg [m-K/W]
borehole thermal resistance total internal resistance is higher by several orders of magnitude. Even though theand second-order andthermal third-order multipole formulas presented in this magnitude. Even though the second-order and third-order multipole formulas presented in this paper are more complicated than other analytical expressions including the zeroth-order and firstpaper are more they complicated than other analytical expressions including the zeroth-order and firstorder formulas, are still simple enough to apply for computation purposes. The implementation order formulas, they are still simple enough to apply for computation purposes. The implementation of the second-order multipole formulas involves approximately ten lines of coding of rather compact of thesimple second-order multipole formulas involves ten of coding of ratherformulas compact and algebraic expressions. Whereas, the approximately implementation oflines third-order multipole Energies 2018, 11, 214 14 of 17 and simple algebraic expressions. Whereas, the implementation of third-order multipole formulas requires programing of a more complex set of equations involving vector and matrix algebra. requires programing of a more complex set of equations involving vector matrix algebra. Although, both second-order and third-order multipole formulas offerand a significant improvement Although, both second-order and third-order multipole formulas offer a significant improvement values, and the right-side ones show the total internal thermal resistance R values, plotted against a over the original implementation of the multipole method, which required about 600 lines of over the original implementation of the multipole method, which required about 600 lines of the grout thermal conductivity. Each figure presents three curves corresponding to close, moderate FORTRAN coding, but the second-order multipole formulas are intrinsically more straightforward FORTRAN coding, but the shown second-order multipole arerespectively. intrinsically more straightforward and wide shank spacings, in black, blue andformulas redaccuracy colors, The exact value of the to implement, while providing almost similar levels of as third-order multipole formulas. to implement, while providing almost similar levels of accuracy as third-order multipole formulas. shank spacing for each case is provided in Table 1. It must be pointed out that multipole formulas Figures 5−7 present a selection of representative comparison results of second-order multiple Figures 5−7 present a selection of representative comparison resultsThe ofand second-order presented in the previous section, calculate the borehole thermal resistance not the groutmultiple thermal formulas against zeroth-order, first-order and tenth-order multipoles. results shown in these formulas against zeroth-order, first-order and tenth-order multipoles. The results shown in these resistance. However, in order to be consistent with the dataset provided by [13], the values of grout figures are for a single ground thermal conductivity of 4.0 W/m-K. The left-side figures show the figures are for a single ground thermal conductivity of 4.0 W/m-K. The left-side figures show the thermal resistance have been calculated and presented in Figures 5–7. The grout thermal resistance grout thermal resistance Rg values, and the right-side ones show the total internal thermal resistance grout thermal resistance Rg values, the right-side show the total internal resistance values have been determined fromand Equation (2), by ones subtracting half of the fixedthermal pipe resistance of R a values, plotted against the grout thermal conductivity. Each figure presents three curves R a values, plotted against the grout thermal conductivity. Each figure presents three curves 0.05 m-K/W from the corresponding borehole thermal resistance values obtained from the multipole corresponding to close, moderate and wide shank spacings, shown in black, blue and red colors, corresponding to exact close,the moderate wide shankdirectly spacings, shown in black, and colors, formulas. Computing grout resistance disregarding theinblue pipe resistance gives respectively. The value of thermal theand shank spacing for eachby case is provided Table 1. red It must be respectively. The exact value of the shank spacing for each case is provided in Table 1. It must be erroneous results for all but zeroth-order multipole calculations. pointed out that multipole formulas presented in the previous section, calculate the borehole thermal pointed outand that multipole formulas inHowever, the previous section, the borehole Figures 5–7 further demonstrate the ability of second-order multiple formulas to calculate the resistance not the grout thermalpresented resistance. in order to calculate be consistent with thethermal dataset resistance and not the grout thermal resistance. However, in order to be consistent with the dataset grout thermal resistance R and the total internal thermal resistance R . The thermal resistance values g a calculated and presented in provided by [13], the values of grout thermal resistance have been provided by [13], the values of grout thermal resistance have been calculated and presented in calculated from the second-order multiple formulas are always within 1% of the original tenth-order Figures 5−7. The grout thermal resistance values have been determined from Equation (2), by Figures 5−7. The grout thermal resistance values have been determined from Equation (2), by multipole method over the entire range of parameters. Hence, it can be concluded that due to their subtracting half of the fixed pipe resistance of 0.05 m-K/W from the corresponding borehole thermal subtracting half ofobtained the fixed pipe ease resistance of 0.05 m-K/W the from the corresponding borehole thermal excellent accuracy and relative of implementation, second-order multipole formulas are resistance values from the multipole formulas. Computing the grout thermal resistance resistance values obtained from the multipole formulas. Computing the grout thermal resistance recommended for calculation of borehole thermal resistance and total internal thermal resistance for directly by disregarding the pipe resistance gives erroneous results for all but zeroth-order multipole directly by disregarding the pipe resistance gives erroneous results for all but zeroth-order multipole all cases where the two legs of the U-tube are placed symmetrically in the borehole. calculations. calculations.
4.2 4.2
0.75 0.75 0.60 0.60 0.45 0.45 0.30 0.30 0.15 0.15 0.00 0.00 0.0 0.0
0th-order Multipole 1st-order Multipole 0th-order Multipole 2nd-orderMultipole Multipole 1st-order 10th-orderMultipole Multipole 2nd-order 10th-order Multipole
0.6 0.6
1.2 1.2
1.8 2.4 2.4 λ1.8 [W/m-K] g λg [W/m-K]
3.0 3.0
3.6 3.6
4.2 4.2
0.40 0.40 0.32 0.32 0.24 0.24 0.16 0.16 0.08 0.08 0.00 0.00 0.0 0.0
0th-order Multipole 0th-order Multipole 1st-order Multipole 2nd-orderMultipole Multipole 1st-order 10th-orderMultipole Multipole 2nd-order 10th-order Multipole
0.6 0.6
1.2 1.2
1.8 2.4 2.4 λ1.8 [W/m-K] g λg [W/m-K]
3.0 3.0
3.6 3.6
4.2 4.2
Ra [m-K/W] Ra [m-K/W]
Rg [m-K/W] Rg [m-K/W]
Figure 5. Grout thermal resistance (Rg) and total internal resistance (Ra) for close (2xp = 32 mm), Figure 5. 5. Grout Grout thermal thermal resistance resistance (Rgg) ) and and total total internal internal resistance resistance (R (Raa)) for for close close (2x (2xpp ==32 32 mm), mm), Figure mm) and wide (2x(R p = 64 mm) configurations with 2rb = 96 mm and λ = 4 W/m-K. moderate (2xp = 43 moderate (2x = 43 mm) and wide (2x = 64 mm) configurations with 2r = 96 mm and λ = 4 W/m-K. moderate (2xpp= 43 mm) and wide (2xpp= 64 mm) configurations with 2rbb= 96 mm and λ = 4 W/m-K. 1.25 1.25 1.00 1.00 0.75 0.75 0.50 0.50 0.25 0.25 0.00 0.00 0.0 0.0
0th-order Multipole 1st-order Multipole 0th-order Multipole 2nd-orderMultipole Multipole 1st-order 10th-orderMultipole Multipole 2nd-order 10th-order Multipole
0.6 0.6
1.2 1.2
1.8 2.4 2.4 λ1.8 [W/m-K] g λg [W/m-K]
3.0 3.0
3.6 3.6
Figure 6. Grout thermal resistance (Rg ) and total internal resistance (Ra ) for close (2xp = 32 mm), moderate (2xp = 75 mm) and wide (2xp = 160 mm) configurations with 2rb = 192 mm and λ = 4 W/m-K.
4.2 4.2
grout thermal resistance Rg and the total internal thermal resistance Ra. The thermal resistance values calculated from the second-order multiple formulas are always within 1% of the original tenth-order multipole method over the entire range of parameters. Hence, it can be concluded that due to their excellent accuracy and relative ease of implementation, the second-order multipole formulas are recommended for calculation of borehole thermal resistance and total internal thermal resistance Energies 2018, 11, 214 15 offor 17 all cases where the two legs of the U-tube are placed symmetrically in the borehole. 0th-order Multipole 1st-order Multipole 2nd-order Multipole 10th-order Multipole
Rg [m-K/W]
0.40 0.30 0.20 0.10
1.50
0th-order Multipole 1st-order Multipole 2nd-order Multipole 10th-order Multipole
1.20
Ra [m-K/W]
0.50
0.90 0.60 0.30
0.00
0.00 0.0
0.6
1.2
1.8
2.4
3.0
3.6
4.2
λg [W/m-K]
0.0
0.6
1.2
1.8
2.4
3.0
3.6
4.2
λg [W/m-K]
Figure 7. Grout thermal resistance (Rg) and total internal resistance (Ra) for close (2xp = 32 mm), Figure 7. Grout thermal resistance (Rg ) and total internal resistance (Ra ) for close (2xp = 32 mm), moderate (2xp = 107 mm) and wide (2xp = 256 mm) configurations with 2rb = 288 mm and moderate (2xp = 107 mm) and wide (2xp = 256 mm) configurations with 2rb = 288 mm and λ = 4 λ = 4 W/m-K. W/m-K.
6. Conclusions 6. Conclusions This paper has presented new second-order and third-order multipole formulas for calculating This paper has presented new second-order and third-order multipole formulas for calculating borehole thermal resistance and total internal thermal resistance. The newly derived formulas can be borehole thermal resistance and total internal thermal resistance. The newly derived formulas can used for all single U-tube applications where the two legs of the U-tube are symmetrically placed in be used for all single U-tube applications where the two legs of the U-tube are symmetrically placed the borehole. The presented formulas have been compared with the original multipole method, as in the borehole. The presented formulas have been compared with the original multipole method, well as the previously derived zeroth-order and first-order explicit multipole formulas. Both the as well as the previously derived zeroth-order and first-order explicit multipole formulas. Both the second-order and third-order multipole formulas provide significant accuracy improvements over second-order and third-order multipole formulas provide significant accuracy improvements over the zeroth-order and the first-order multipole formulations. The thermal resistances calculated from the zeroth-order and the first-order multipole formulations. The thermal resistances calculated from the second-order and third-order multipole formulas are always within 1% and 0.2% of the original the second-order and third-order multipole formulas are always within 1% and 0.2% of the original tenth-order multipole method, respectively. Second-order multipole formulas are, however, more tenth-order multipole method, respectively. Second-order multipole formulas are, however, more straightforward to use and are, thus, recommended for use in practice. straightforward to use and are, thus, recommended for use in practice. Acknowledgments: Thesupport supportofofboth bothauthors authorswas wasprovided providedby bythe theSwedish SwedishEnergy EnergyAgency Agency(Energimyndigheten) (Energimyndigheten) Acknowledgments:The through its national nationalresearch research program EFFSYS Expand. funding source had no involvement in thedesign; study through its program EFFSYS Expand. TheThe funding source had no involvement in the study collection, analysis analysis and interpretation of data; orofindata; the decision submit the article for design; collection, and interpretation or in thetodecision to submit thepublication. article for publication. Author Contributions: Contributions:Both Bothauthors authorscontributed contributedequally equallyin inthe thedesign, design,concept, concept,and andpreparation preparationof ofthe themanuscript. manuscript. Author Conflicts of The authors authors declare Conflicts of Interest: Interest: The declare no no conflict conflict of of interest. interest.
Nomenclature Nomenclature B± J± bk bk cf cf H H J J N N Pn,j P±n,j Pn,j p0 p1 p2 qb qn Ra Rb
= = = = = = = = = = = = = = = = = = = = = =
Multipole correction for multipole order J for even (+) and odd (−) cases, See Equation (30) Multipole correction for multipole order J for even (+) and odd (−) cases, See Equation (30) Dimensionless parameter, see Equation (29) Dimensionless parameter, see Equation (29) Specific heat of the circulating fluid in the U-tube, J/kg-K Specific heat of the circulating fluid in the U-tube, J/kg-K Depth of the borehole, m Depth of the borehole, m Number of multipoles, j = 0, 1, 2, . . . , J Number of multipoles, j = 0, 1, 2 … J Number of pipes in the borehole; N = 2 for single U-tube Number of pipes in the borehole; N = 2 for single U-tube Multipole factor of order j for pipe n Multipole factor of order j for pipe n Multipole factor of order j for pipe n for even (+) and odd (−) cases Dimensionless parameter, see Equation (29) Dimensionless parameter, see Equation (29) Dimensionless parameter, see Equation (29) Heat rejection rate per unit length of borehole, W/m Heat rejection rate per unit length of pipe n, W/m Total internal borehole thermal resistance, m-K/W Local borehole thermal resistance between fluid in the U-tube to borehole wall, m-K/W
Energies 2018, 11, 214
Rb∗ Rg R± J Rn b Rp R12 rb rp Tb Tb,av Tf Tf,loc Tfn ± Tfn Vf xp β λ λb ρf σ
= = = = = = = = = = = = = = = = = = = = =
16 of 17
Effective borehole thermal resistance, m-K/W Grout thermal resistance, m-K/W Thermal resistance for even (+) and odd (−) cases for J multipoles, m-K/W Thermal resistance between U-tube leg n and borehole wall, m-K/W Total fluid-to-pipe resistance for a single pipe i.e., one leg of the U-tube, m-K/W Thermal resistance between U-tube legs 1 and 2, m-K/W Radius of the borehole, m Outer radius of the pipe making up the U-tube, m Borehole wall temperature, ◦ C Average temperature at the borehole wall, ◦ C Mean fluid temperature inside the U-tube, ◦ C Local mean fluid temperature, ◦ C Fluid temperature in U-tube leg n, ◦ C Fluid temperature in U-tube leg n for even and odd cases, ◦ C Volume flow rate of the circulating fluid in the U-tube, m3 /s Half shank spacing; half of center-to-center distance between U-tube legs, m, See Figure 3 Dimensionless thermal resistance of one U-tube leg, see Equation (20) Thermal conductivity of the ground, W/m-K Thermal conductivity of the borehole/grout, W/m-K Density of the circulating fluid in the U-tube, kg/m3 Thermal conductivity ratio, dimensionless, see Equation (20)
References 1. 2.
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.
Lund, J.W.; Boyd, T.L. Direct utilization of geothermal energy 2015 worldwide review. Geothermics 2016, 60, 66–93. [CrossRef] Vieira, A.; Alberdi-Pagola, M.; Christodoulides, P.; Javed, S.; Loveridge, F.; Nguyen, F.; Cecinato, F.; Maranha, J.; Florides, G.; Prodan, I.; et al. Characterisation of ground thermal and thermo-mechanical behaviour for shallow geothermal energy applications. Energies 2017, 10, 2044. [CrossRef] Javed, S.; Spitler, J.D. Calculation of borehole thermal resistance. In Advances in Ground-Source Heat Pump Systems, 1st ed.; Rees, S.J., Ed.; Woodhead Publishing: Cambridge, UK, 2016; pp. 63–95. ISBN 978-0-08-100311-4. Spitler, J.D.; Gehlin, S.E. Thermal response testing for ground source heat pump systems—An historical review. Renew. Sustain. Energy Rev. 2015, 50, 1125–1137. [CrossRef] Kavanaugh, S.P. Simulation and Experimental Verification of Vertical Ground-Coupled Heat Pump Systems. Ph.D. Thesis, Oklahoma State University, Stillwater, OK, USA, May 1985. Gu, Y.; O’Neal, D.L. Development of an equivalent diameter expression for vertical U-tubes used in ground-coupled heat pumps. ASHRAE Trans. 1998, 104, 347–355. Shonder, J.A.; Beck, J.V. Field test of a new method for determining soil formation thermal conductivity and borehole resistance. ASHRAE Trans. 2000, 106, 843–850. Paul, N.D. The Effect of Grout Thermal Conductivity on Vertical Geothermal Heat Exchanger Design and Performance. Master’s Thesis, South Dakota University, Brookings, SD, USA, July 1996. Sharqawy, M.H.; Mokheimer, E.M.; Badr, H.M. Effective pipe-to-borehole thermal resistance for vertical ground heat exchangers. Geothermics 2009, 38, 271–277. [CrossRef] Bauer, D.; Heidemann, W.; Müller-Steinhagen, H.; Diersch, H.J. Thermal resistance and capacity models for borehole heat exchangers. Int. J. Energy Res. 2011, 35, 312–320. [CrossRef] Liao, Q.; Zhou, C.; Cui, W.; Jen, T.C. New correlations for thermal resistances of vertical single U-Tube ground heat exchanger. J. Therm. Sci. Eng. Appl. 2012, 4, 031010. [CrossRef] Lamarche, L.; Kajl, S.; Beauchamp, B. A review of methods to evaluate borehole thermal resistances in geothermal heat-pump systems. Geothermics 2010, 39, 187–200. [CrossRef] Javed, S.; Spitler, J.D. Accuracy of borehole thermal resistance calculation methods for grouted single U-tube ground heat exchangers. Appl. Energy 2017, 187, 790–806. [CrossRef] Bennet, J.; Claesson, J.; Hellström, G. Multipole Method to Compute the Conductive Heat Flows to and between Pipes in a Composite Cylinder. Notes on Heat Transfer 3; University of Lund: Lund, Sweden, 1987.
Energies 2018, 11, 214
15. 16. 17.
18. 19. 20. 21.
22. 23. 24. 25. 26.
17 of 17
Claesson, J.; Bennet, J. Multipole Method to Compute the Conductive Heat Flows to and between Pipes in a Cylinder. Notes on Heat Transfer 2; University of Lund: Lund, Sweden, 1987. Claesson, J.; Hellström, G. Multipole method to calculate borehole thermal resistances in a borehole heat exchanger. HVAC R Res. 2011, 17, 895–911. Young, T.R. Development, Verification, and Design Analysis of the Borehole Fluid Thermal Mass Model for Approximating Short Term Borehole Thermal Response. Ph.D. Thesis, Oklahoma State University, Stillwater, OK, USA, December 2004. He, M. Numerical Modelling of Geothermal Borehole Heat Exchanger Systems. Ph.D. Thesis, De Montfort University, Leicester, UK, February 2012. Al-Chalabi, R. Thermal Resistance of U-Tube Borehole Heat Exchanger System: Numerical Study. Master’s Thesis, University of Manchester, Manchester, UK, September 2013. Earth Energy Designer (EED), v3.2; BLOCON: Lund, Sweden, 2015. Spitler, J.D. GLHEPRO—A Design Tool for Commercial Building Ground Loop Heat Exchangers. In Proceedings of the Fourth International Heat Pumps in Cold Climates Conference, Aylmer, QC, Canada, 17–18 August 2000. Claesson, J. Multipole Method to Calculate Borehole Thermal Resistances; Mathematical Report; Chalmers University of Technology: Gothenburg, Sweden, 2012. Hellström, G. Ground Heat Storage—Thermal Analyses of Duct Storage Systems—Theory. Ph.D. Thesis, University of Lund, Lund, Sweden, April 1991. Spitler, J.D.; Javed, S.; Ramstad, R.K. Natural convection in groundwater-filled boreholes used as ground heat exchangers. Appl. Energy 2016, 164, 352–365. [CrossRef] Zeng, H.; Diao, N.; Fang, Z. Heat transfer analysis of boreholes in vertical ground heat exchangers. Int. J. Heat Mass Transf. 2003, 46, 4467–4481. [CrossRef] Ma, W.; Li, M.; Li, P.; Lai, A.C. New quasi-3D model for heat transfer in U-shaped GHEs (ground heat exchangers): Effective overall thermal resistance. Energy 2015, 90, 578–587. [CrossRef] © 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
